We have shown that over-expression of a wild type human transthyretin (TTR) gene can suppress the neuropathologic and behavioral phenotypes of a well-studied validated transgenic model of human Alzheimer's disease (AD). We have also shown that TTR interacts with A2 in vitro to inhibit oligomer and fibril formation and tissue culture cytotoxicity and in vivo with TTR-A2 complexes being co-immunoprecipitable from APP23 and some human AD brains with anti-sera to either protein. We have also noted that 70% of human cortical neurons from AD brains stain with an anti-TTR antibody (compared with 10% of age matched control brains) as do almost all the hippocampal and cortical neurons in the APP23 mouse model. TTR mRNA (by qtPCR) is 10 times higher in primary hippocampal neurons from APP23 mice than from controls and the ratio is even greater in hippocampal lysates. These findings coupled with the observation that the pathologic changes appear to be accelerated when the APP23 strain is crossed onto a mouse TTR knockout background suggest that TTR may be a normal physiologic neuronal defense mechanism in neurodegenerative disease of the AD type. In contrast the suppression of the features of AD by TTR over-expression may be viewed as pharmacologic. Our proposal is directed at finding small molecules that will enhance TTR transcription and protein production in cells of neuronal origin. To that end we have developed reporter constructs in which the human TTR promoter drives a Gaussia luciferase reporter gene which is readily detectable in a high throughput screening mode available at our institution. Pilot studies have shown that the assay is robust, reproducible and stable under a variety of conditions. We propose to use the assay to serially screen a group of available molecular libraries (Maybridge, LOPAC, MLPCN, Scripps Florida) to identify the 25 most potent, specific least toxic enhancers of TTR transcription in neuronally defined cells. We will characterize their mechanism of action using cell biologic, proteomic and microarray techniques established in our laboratory and ultimately assay their capacity to suppress the neuropathologic and behavioral manifestations displayed by the APP23 transgenic mouse model of AD. PUBLIC HEALTH RELEVANCE: We have been able to suppress the appearance of the characteristic plaques and the associated learning difficulties seen in a well-validated mouse model of Alzheimer's disease by genetically inserting many copies of an apparently beneficial gene into the AD mice. We now wish to find compounds that can be administered to the mice (and ultimately humans) that will increase the production of the beneficial protein in neurons thus rendering them resistant to the toxic effects of the AD protein.